Evaluating the quality of a similar pages suggestion tool in Dutch [NL]

We developed a tool that provides similar pages suggestion for a given page of a website.

Something really similar to what Amazon does on its websites: when you browse the page of a product (i.e. "iphone 13") it provides you links to similar product pages inside a "similar products" widget (i.e. "iphone 12", "iphone 11", "iphone 13 cases", etc...).

We need someone that help us to evaluate the quality of such pages suggestion tool.

You'll be provided with an xlsx file containing a list of 2.000 rows of input text (i.e. "iphone 13", "bird cage", "bmw series 3") and for each one of them a list of results from our tool (i.e. for the input text "iphone 13" you'll have the following list of 10 elements ["iphone 13", "iphone 11", "iphone 13 cases", "apple", "macbook pro", "macbook pro 13", "ipad", "ipad 3"].).

Your job will be to score the similarity of each row of elements with a number from 1 to 4 where 1 = super low similarity, 2 = low similarity, 3 = good similarity and 4 = super good similarity.

Some examples:

Input text: "iphone" - Result set ["iphone 13", "iphone 11", "iphone 13 cases", "apple", "macbook pro", "macbook pro 13", "ipad", "ipad 3"] would be scored as 4 = super good similarity

Input text: "iphone" - Result set ["used cars", "bicycles", "bird cages", "dog food", "cat food"] would be scored as 1 = super low similarity

You'll receive additional details about cases and ratings in the excel file.

In this case all the text will be in dutch (NL) so we're searching for someone who's dutch mother tongue only.

We estimate something like 5/6 hours of work to be completed (2.000 rows x 10 seconds for evaluation of each one of them).

In the attachment you can see an example of the file that will be provided to you.